addendum to repeated counting of multiple response category of 'none'

[J McClure]

Hi,
When I extracted the information from the records, if a respondent had no psychiatric diagnoses I recorded a 'none' for variables 1 through 6. So, if I recode the responses from 'none' to 'missing' for variables 2 through 6 it will solve my problem since the only 'none' response that will be counted is for the first variable. Does this sound right?
Thanks,
Jan

-------- Original Message --------

Subject:	repeated counting of multiple response category of 'none'
Date:	Sat, 28 Aug 2010 08:10:37 -0700
From:	J McClure [hidden email]
To:	[hidden email]

Hi,
I have a multiple response situation. I extracted from medical records
the respondent's psychiatric diagnoses. Some respondents had no
diagnoses, some 1, some 2 etc. I created 6 variables (Psy1 to Psy6) to
record the results. Each variable has 8 possible values coded 0 to 7.
The first possible value is '0' or none.  My problem is that in the
resulting crosstabs table, 'none' gets counted 1813 times for 430
respondents since almost everyone has fewer than 6 diagnoses and thus
comes up with a none in one or more of the variables. I want to count
'none' only one time for an individual respondent.
Thanks, Jan
I used the below syntax I found on the listserv:
AUTORECODE VARIABLES=Psy1 Psy2 Psy3 Psy4  Psy5 Psy6
 /INTO nPsy1 nPsy2 nPsy3 nPsy4  nPsy5 nPsy6
 /GROUP
 /BLANK=MISSING
 /PRINT.
MULT RESPONSE
 /groups=diagnoses 'diagnoses' ((1,8))
 /variables = MR_sex (0,1)
 /tables= MR_sex by diagnoses
 /base = cases
 /cells = all.

===================== To manage your subscription to SPSSX-L, send a message to [hidden email] (not to SPSSX-L), with no body text except the command. To leave the list, send the command SIGNOFF SPSSX-L For a list of commands to manage subscriptions, send the command INFO REFCARD